AND/OR Importance Sampling

Abstract

The paper introduces AND/OR importance sampling for probabilistic graphical models. In contrast to importance sampling, AND/OR importance sampling caches samples in the AND/OR space and then extracts a new sample mean from the stored samples. We prove that AND/OR importance sampling may have lower variance than importance sampling; thereby providing a theoretical justification for preferring it over importance sampling. Our empirical evaluation demonstrates that AND/OR importance sampling is far more accurate than importance sampling in many cases.

Cite

Text

Gogate and Dechter. "AND/OR Importance Sampling." Conference on Uncertainty in Artificial Intelligence, 2008.

Markdown

[Gogate and Dechter. "AND/OR Importance Sampling." Conference on Uncertainty in Artificial Intelligence, 2008.](https://mlanthology.org/uai/2008/gogate2008uai-importance/)

BibTeX

@inproceedings{gogate2008uai-importance,
  title     = {{AND/OR Importance Sampling}},
  author    = {Gogate, Vibhav and Dechter, Rina},
  booktitle = {Conference on Uncertainty in Artificial Intelligence},
  year      = {2008},
  pages     = {212-219},
  url       = {https://mlanthology.org/uai/2008/gogate2008uai-importance/}
}